Principles of Neurosurgery,

edited by Robert G. Grossman. Rosenberg © 1991.
Published by Raven Press, Ltd., New York.

CHAPTER 14

Movement Disorders

Robert G. Grossman and Winifred J. Hamilton

Introduction, 305
Stereotactic Thalamic Lesioning, 305

Neurophysiological Basis, 305
Indications, 307
Stereotactic Technique and Considerations, 310

Other Procedures, 314

Nerve-Sectioning Procedures, 314
Chronic Electrical Stimulation, 315
Neurotransmitter Augmentation, 315

References, 316

INTRODUCTION

A number of movement disorders can be ameliorated,
sometimes even abolished, by neurosurgical procedures.
The most commonly performed procedure is the stereo-
tactic placement of a lesion in the ventrolateral thala-
mus. Such a lesion usually abolishes the tremor of Par-
kinson's disease or essential tremor, and it can reduce
the rigidity associated with Parkinson's disease (1,2). A
single ventrolateral thalamic lesion is less successful for
treating hemiballismus, dystonia, or torticollis, although
in individual cases, it sometimes produces significant re-
lief. More extensive thalamic lesioning, or selective
nerve and muscle sectioning in the case of torticollis, can
sometimes lessen these dyskinesias although complete
abolition is generally not possible. Several other neuro-
surgical procedures have been attempted for particular
dyskinesias, with varying degrees of success. These in-
clude chronic electrical stimulation, implanted pumps
for the chronic release of neurotransmitters or medica-
tions, and the transplantation of neural tissue.

STEREOTACTIC THALAMIC LESIONING

The use of thalamic lesioning in the treatment of

movement disorders has largely developed through the
study and treatment of Parkinson's disease. Parkinson's
disease is the most prevalent and the most extensively
studied of the movement disorders. It is estimated that in

R. G. Grossman and W. J. Hamilton: Department of Neuro-

surgery, Baylor College of Medicine, Houston, Texas 77030.

the United States alone there are 40,000 new cases each
year and a cumulative total of approximately 300,000
cases at any one time (3,4).

Neurophysiological Basis

Any attempt to understand the efficacy of a thalamic

lesion in ameliorating the tremor of Parkinson's disease
must grapple with a divergent array of physiological, ana-
tomical, and clinical evidence; this evidence, although it
suggests certain models by which abnormal movements
may be generated, also reveals how little is understood
about the pathophysiology underlying the movement
disorders. The primary pathology observed in Parkin-
son's disease is the deterioration of dopaminergic neu-
rons in the midbrain, especially in the substantia nigra,
and a concomitant deterioration of the dopaminergic
pathways projecting to the striatum (5-7). The etiology
of this deterioration is not known. Dopaminergic projec-
tions from the midbrain have an inhibitory influence on
the striatum, and it has therefore been suggested that the
loss of this influence may disinhibit the striatum and
thereby lead to tremor (8). Also, dopamine inhibits stria-
tal production of acetylcholine; therefore, loss of dopa-
mine leads to excess acetylcholine, which presumably
contributes to the symptoms of Parkinson's disease.

There is evidence to suggest that the mechanisms un-

derlying tremor and bradykinesia are fundamentally dif-
ferent. Surgery is very effective for tremor, but it has little
or no effect on bradykinesia. Conversely, levodopa or
carbidopa is effective for bradykinesia but has less effect
on tremor. Hoehn and associates found decreased levels

305

306 / CHAPTER 14

of dopamine catabolites in the urine of patients with Par-

kinson's disease and noted that these levels correlated

with the degree of bradykinesia, rigidity, and mental dete-

rioration observed. They found no correlation between

the levels of dopamine catabolites and tremor (9).

The complexity of the pathophysiological substrate of

Parkinson's disease is underscored by the fact that

tremor can be relieved to a greater or lesser degree by a
lesion in several cortical and subcortical areas—includ-
ing cortical areas 4 or 6, the globus pallidus, the ventro-
lateral thalamus, and the ansa and fasicularis lenticularis
(10-12). Of these sites, most surgeons now prefer placing
a lesion in the ventralis intermedius (VIM) of the ventro-
lateral thalamus (according to Hasslcr's terminology; see
Fig. 1). Such a lesion is effective in ameliorating tremor
and creates few, if any, secondary neurological deficits

(1,2,13). In addition, a VIM lesion ameliorates rigidity in

some parkinsonian patients (1,2), although a slightly
more anterior lesion, in the ventralis oralis posterior

(VOP) or the ventralis oralis anterior (VOA), is often

advocated for the treatment of patients in whom rigidity
predominates (5,8,14). Although long-lasting relief from
the tremor of Parkinson's disease has been reported fol-

lowing a VIM lesion (1,2), the lesion does not affect the

progression of bradykinesia.

Although it is unclear how a VIM lesion alters the

pathophysiology that leads to the tremor of Parkinson's
disease, several aspects of the neurophysiology of the

ventrolateral thalamus pertain to this question. First,
unit recordings made in the VIM nucleus of the thala-
mus have revealed rhythmic bursts that are synchronous

RG. 1. Division of the ventrolateral thalamus according to

the terminology established by Hassler. Abbreviations: CM =

centrum medianum; Gpe = external segment of globus palli-
dus; Gpi = internal segment of globus pallidus; LPO = lateral
polaris; Put = putamen; Vce = ventralis caudalis externus;
Vci = ventraiis caudalis internus; VIM = ventralis intermedius;
VOA = ventralis oralis anterior; VOP = ventralis oralis poste-

rior. (From reference 17, with permission.)

with the tremor in patients with Parkinson's disease (15).
The importance of this area in the generation of tremor
is further supported by the fact that stimulation in the
VIM nucleus will alter the tremor, either driving the

tremor or, less commonly, suppressing it (14,15). Some

researchers have suggested that the neurons of the VIM
nucleus may autogenously generate the discharge and

that therefore destruction of these neurons stops the
tremor, but it is also possible, for the following reasons,

that the lesion interrupts and alters the complex neural

circuitry by which movements are monitored and regu-
lated.

Lesions made in the ventrolateral thalamus that are

effective in relieving tremor and rigidity appear to in-

terrupt two pathways within the subcortical circuitry,

which eventually drives cortical firing (Fig. 2) (5,16,17).
One is the pallidothalamic pathway extending from the
globus pallidus via the ansa lenticularis of the fields of
Forel to the anterior-inferior aspect (the area of the
VOA) of the ventrolateral thalamus and on to the cortex,

particularly area 6. The other pathway includes fibers

from the contralateral cerebellum that pass through the
thalamic peduncle into the more posterior parts of the

ventrolateral thalamus (VIM-VOP) and continue on to
area 4 of the cortex. It has been suggested that the first

pathway may be associated primarily with rigidity,
whereas the second may be associated primarily with
tremor (17).

Hassler has proposed that depression of the fusimotor

(gamma) system may underlie the tremor and rigidity of
Parkinson's disease and that a ventrolateral thalamic le-
sion, by reducing thalamic inhibition of the fusimotor
system, may help to restore proper fusimotor activity
(17). In support of this general hypothesis, Stern and
Ward produced a parkinsonian-like state in the monkey
in two ways: by administering reserpine and by making
lesions in the ventromedial tegmentum of the mesen-

cephalon (18,19). Both models produce decreased levels

of dopamine in the striatum and a decrease in gamma
efferent activity with resultant alpha phasic and tonic
hyperactivity. They postulate that the experimental le-

FIG. 2. Schematic presentation of the major interconnec-
tions between the motor cortex and the subcortical motor
areas. (From reference 27, with permission.)

Sensory

Systems

Motor

Cortex

Corticospinal

Tract

Caudate Nucleus

and Putamen

Substantia Nigra

Midbrain

Globus Pallidus

Ventrolateral

Thalamus

Cerebellum

Subthalamic

Nucleus

MOVEMENT DISORDERS / 307

sions disturbed the balance between the alpha and

gamma motor neurons, giving rise to a preponderance of

alpha tone. Similarly, they argue, Parkinson's disease
may alter the balance between the alpha and gamma
motor systems, and a thalamic lesion may help to restore
the normal balance (20). Further support for the theory
of gamma inhibition is provided by the fact that chlor-
promazine, in appropriate doses, inhibits the gamma
system. Syndromes resembling Parkinson's disease are

produced in some patients overtreated with chlorproma-
zine(21).

However, other studies have demonstrated a height-

ened, rather than a depressed, level of fusimotor activity
in patients with Parkinson's disease (12). This finding is

supported by studies that have revealed a change in the

mode of maintained voluntary innervation after a ven-
trolateral thalamic lesion, suggesting a shift from alpha-

gamma coactivation to a predominantly alpha-driven

muscle innervation (22). Yanagisawa and associates

found that bilateral ablation of the sensory-motor cor-

tex, cerebellum, and medullary pyramids failed to re-
duce the muscle spindle's response to thalamic stimula-
tion, whereas bilateral lesions in the midbrain

tegmentum did. These findings suggest that the facilita-
tion is mediated by a pathway from the ventrolateral
thalamus to the reticular formation (23).

Lieberman, Copack, and Oilman noted a decrease in

fusimotor activity leading to a decrease in spindle affer-
ent responses following a ventrolateral thalamic lesion
(24). This decrease in response deprives the alpha neu-
rons of a strong facilitatory influence and, they postu-

late, would result in a decrease of previously hyperactive
alpha activity and a concomitant reduction in muscle
tension. Their findings support those of Yanagisawa and
associates by suggesting that the ventrolateral thalamus
exerts a net excitatory influence on the fusimotor neu-

rons, rather than a net inhibitory effect as proposed by

Hassler and by Stern and Ward.

Indications

Parkinson's Disease

Between 1950 and 1968, surgical neuroablative proce-

dures played a major role in the treatment of Parkinson's

disease. With the introduction of levodopa in 1968, how-
ever, utilization of surgical treatment substantially de-
clined (25). More recently, improvements in stereotactic
surgery and a growing understanding of the limitations

and side effects of levodopa therapy have renewed inter-

est in the use of surgery for patients with Parkinson's

disease (5,26).

Although medical control of symptoms is generally

attempted first, some clinicians believe that surgery
should be used to treat the early-appearing symptoms of

tremor and rigidity in relatively young patients. This
way, levodopa can be reserved for the usually later-de-
veloping and more debilitating symptom of bradykine-
sia, for which there is presently no other effective treat-
ment (1,5). Studies have also found that patients with a
pre-existing thalamotomy suffer fewer side effects from
levodopa therapy. Not only do these patients appear to
have fewer levodopa-induced dyskinesias in the limbs

contralateral to the lesion (1), but thalamotomy appears

to lessen the "on-off" syndrome associated with contin-

ued levodopa therapy (5). Narabayashi and associates
noted that levodopa-induced dyskinesias were largely
eliminated by a lesion in VIM-VOP (14). Nagaseki and

coworkers reported that, in a series of 27 patients oper-

ated on for Parkinson's disease, the amount of levodopa

necessary to control symptoms after thalamotomy was

approximately one-half of that necessary before sur-
gery (2).

The ideal candidate for surgical treatment of Parkin-

son's disease is a patient who (1) has unilateral tremor

and little or no rigidity, bradykinesia, mental deteriora-

tion, or speech and gait disturbances; (2) is unresponsive

to or intolerant of drug treatment; and (3) has no general

medical risk factors that would increase the risk of sur-

gery. The decision to offer surgery to less-than-ideal can-
didates—patients with rigidity, bradykinesia. and ill-
nesses common in older patients—should be based on a

careful weighing of the benefits and risks for the individ-

ual patient. Unilateral tremor can be eliminated by a
VIM lesion in 85 to 90 percent of patients and can be

ameliorated in the remainder (1,2). Rigidity appears to

be lessened in some patients following a VIM lesion (2).
However, as noted earlier, a lesion slightly anterior to the

VTM nucleus, in the VOA-VOP complex, is usually
more successful in relieving rigidity than is a VIM lesion

(5,8,15). Kelly and Gillingham found that 88 percent of

58 patients who had significant rigidity preoperatively

had no signs of rigidity two years after VOA thalamo-
tomy (1,8). Most investigators report little or no improve-

ment in bradykinesia following a thalamotomy. Masked
facies, speech and swallowing impairments, festinating

gait, and intellectual impairment also fail to respond to
thalamotomy and may even be aggravated by the sur-
gery (12). Younger patients respond better than older
patients, with older patients experiencing an increased

incidence of postoperative confusion and disturbances

of balance (12,27).

For many patients, a VIM lesion provides long-term,

possibly even permanent, relief from tremor on the side
contralateral to the lesion. Tremor, usually mild and not
disabling, may recur gradually in a percentage of patients
in the years following surgery. In Kelly and Gillingham's
long-term follow-up of 57 patients who underwent thala-
motomy for tremor, 90 percent had no tremor two years
after surgery, 73 percent had no tremor after six years,

and 57 percent had no tremor after 10 years (1). Their

308 CHAPTER 14

findings with regard to rigidity were similar, with 86 per-

cent having no evidence of rigidity on the contralateral
side two years after surgery and 55 percent having no
evidence of rigidity 10 years after surgery. Nagaseki and
associates have also followed patients operated on for
tremor for an extended period (mean 6.58 years). Their
late results were essentially identical to the early results,
with 92 percent of both groups of patients displaying
satisfactory relief (mild or no residual tremor) on the side
contralateral to the surgery (2). The occurrence or pro-
gression of tremor and rigidity on the ipsilateral side was
unaffected by surgery.

For patients with bilateral tremor and rigidity, a sec-

ond thalamic lesion on the opposite side appears to be

less effective than the first lesion and is more apt to be

followed by neurological deficits, especially dysarthria
and disequilibrium (5,12,28). Cooper and associates re-
ported approximately 70 percent relief of tremor and
rigidity following a ventrolateral thalamic lesion on the

second side, in contrast to 90 percent relief following the

first lesion (29). To avoid some of the complications of
bilateral surgery in patients with bilateral Parkinson's
disease, Benadid and coworkers have used a combined

technique of thalamotomy for the most disabled side

and continuous VIM stimulation for the other side, us-
ing stereotactically implanted electrodes connected to
subcutaneous stimulators (30). Stimulation at 130 Hz
strongly decreased the tremor but did not suppress it as
completely as thalamotomy.

Intention Tremor

Tremor activated by movement—intention tremor—

may occur in victims of severe closed head injury and in

patients with multiple sclerosis. Although the tremor is

usually not the most disabling of the neurological deficits
encountered in these two groups of patients, it adds to

the overall disability and often interferes significantly
with activities of daily living. This tremor can be ame-

liorated by a VIM lesion, but complete cessation of the
tremor cannot be expected in most individuals, probably
because of the widespread nature of the subcortical le-
sions in these conditions (31,32). The size of an effective
lesion is generally larger than that required for treatment

of Parkinson's disease.

Bullard and Nashold reported on the results following

a ventrolateral thalamic lesion (in VIM, sometimes ex-
tending into the VOP) in 11 individuals with movement
disorders following a severe head injury (33). Of these,
seven had intention tremor, nine hemiballistic move-
ments, two choreoathetoid movements, and two truncal

ataxia. All 11 had preoperative dysarthria, which ranged

from mild to severe. Postoperatively, all had immediate

improvement in their tremor, although in two the im-
provement was not considered functionally significant.

In four of nine patients available for follow-up examina-
tion, the tremor continued to improve in the interval
since surgery. Five of the six patients who had a left-sided
procedure and one of five patients with a right-sided pro-
cedure had significantly worsened dysarthria following
surgery. In all but two, this dysfunction later improved.

These results generally agree with those of Andrew,

Fowler, and Harrison who found that, following stereo-
tactic thalamotomy, in eight patients with posttraumatic

tremor, all showed improvement (32). The lesion was
designed to be situated in the VIM, close to its boundary
with the ventralis caudalis externus (Vce). In two pa-
tients a second lesion was necessary to stop the tremor,
and in five patients a third lesion was necessary. The

second and third lesions were generally placed more ante-
riorly, in the oralis externus (the external aspect of

Hassler's VOA-VOP). Resting and postural tremors
were abolished in all eight patients, as were myoclonic

jerks. Intention tremor was eliminated in one patient

and greatly reduced in the other seven. All showed

marked functional improvement. Dysarthria and ataxia
tended to be worse immediately after surgery but to
abate in the subsequent weeks.

Cooper has reported lasting abolition of intention

tremor 10 years following surgery in 72 percent of 32
patients with multiple sclerosis (12). In this series, spastic-
ity worsened following surgery in two patients with pre-
existing spasticity, but other symptoms of cerebellar dys-
function (dysarthria, ataxia, hypotonia) were unaffected.
Krayenbiihl and Yasargil have reported similar results
(34). The preferred target appears to be the posterior
aspect of the VIM nucleus, extending just into the Vce

nucleus, although some surgeons advocate a subthala-

mic target (5). Acute exacerbation of the visual, spinal,
and cerebral symptoms of multiple sclerosis has been
reported following thalamotomy; dexamethasone dur-
ing the immediate postoperative period appears to re-
duce this complication (5).

Intention tremor as a sole symptom, so-called "essen-

tial" tremor, can usually be eliminated by a VIM lesion,
although the lesion must generally be larger than for
treatment of Parkinson's disease.

Hemiballismus

Hemiballismus is characterized by violent flinging

motions that involve the extremities on one-half of the

body, although it is usually most marked in the arm.

Most instances occur in the elderly, usually following an
infarction in the subthalmic nucleus, but cases have been
recorded in association with neoplasms, granulomas,
head injury, and multiple sclerosis (33,35). Although rel-
atively few patients have been operated on for hemibal-

lismus, a VIM lesion appears to be effective for 50 to 60

percent of patients with this disorder (12,36). In Bullard
and Nashold's series of 11 patients with dyskinesias sec-

ondary to severe head injury, nine displayed hemiballis-
mic movements. Following thalamotomy, five showed
marked improvement, three moderate improvement,
and one mild improvement (33).

Dystonla

Dystonia is a condition characterized by sustained

muscle contractions that lead to twisting and repetitive

movements and often to abnormal posturing. Its mani-
festation and clinical course are highly variable, suggest-
ing that various etiologies may be involved. The best
surgical results have been obtained in posttraumatic
hemidystonia in which CT or MR images show a struc-

tural lesion, such as infarction in the basal ganglia (37).

The ameliorating effect is greatest for dystonic move-
ments of the hand, less for leg movements, and least for

dystonia of the trunk. Interestingly, the effect of surgery
for dystonia tends to increase over one to three months
following surgery, as opposed to the immediate benefit
observed in patients with parkinsonian tremor. Multiple
thalamic lesions covering several nuclei are generally nec-

essary to ameliorate dystonia. In addition to a lesion in

the VIM nucleus, a lesion is often placed in the centrum

medianum (CM) nucleus (37). particularly if the dys-

tonia is driven by arousal or sensory activation. In other
patients, lesions in the VOP and VOA of the thalamus

may be necessary.

Torticollis

Torticollis differs from the other movement disorders

in that it is confined to the muscles of the head and neck
—primarily the sternocleidomastoid, trapezius, and
splenius capitus muscles, although other muscles may be
involved as well. It is characterized by sustained muscle
contractions that commonly force the head into an ab-
normal posture, usually displaying some degree of rota-
tion and retroflexion, anteflexion, lateral flexion, or any
combination of these. Although torticollis may be a form
of dystonia confined to the head and neck, this is not
clear. Torticollis is, in general, difficult to ameliorate by
medical or surgical means. Recently, the injection of bot-

ulinum toxin into the affected muscles has produced im-
provement in some patients (38). Thalamotomy is gener-
ally not indicated for torticollis except in patients in
whom it is a clear symptom of more generalized dystonia
and in whom other approaches have not provided suffi-
cient relief. For most patients with medically intractable

torticollis, a nerve-sectioning procedure that denervates

the anterior roots of the upper cervical nerves and some-
times the spinal accessary nerve—and in some cases a

combination of this procedure and selected myotomies
—can provide significant relief (5,27,39,40). These pro-

MOVEMENT DISORDERS / 309

cedures for the treatment of torticollis are discussed to-
ward the end of this chapter.

In patients in whom thalamotomy has been per-

formed for torticollis, the results have been variable.
Long-term relief generally requires bilateral lesions,
which increase the risk of dysarthria postoperatively. Sev-

eral studies have shown a permanent speech and swal-
lowing impairment in approximately 5 percent of pa-
tients following surgery (12,32,36). Recurrence of
torticollis several months postoperatively is relatively
common. The ideal target is not clear. In Cooper's series,
bilateral lesions were placed in the medial side of the
ventrolateral thalamus (12); Mundinger and coworkers
have advocated a lesion in the zona incerta, just beneath

the thalamus (36); and Bertrand, Molina-Negro, and

Martinez prefer a point in the ventrolateral thalamus
adjacent to the internal capsule (39).

Choreoathetosis

Athetosis is characterized by slow, repetitive writhing

movements, whereas choreic movements are sudden, ir-
regular, and relatively rapid. This complex of symptoms
is most commonly seen in patients with cerebral palsy.
These patients generally suffer from other types of move-
ment disorders as well, including dystonia. ataxia, and
spasticity. Surgery for the symptoms of cerebral palsy
has included ventrolateral thalamotomy (for dystonic
and choreoathetotic movements), lesions of the dentate
nucleus of the cerebellum (for choreoathetotic move-
ments), chronic electrical stimulation of the anterior cer-

ebellum or cervical cord (for spasticity), and lumbar sec-
tioning of nerves (for spasticity).

The results of surgery are variable. Because of the wide

variability in the pathology underlying cerebral palsy, it
is difficult to generalize from one patient to another.
Also, because of the severity of the movement disorders
in many of these patients, surgery must often be done
under general anesthesia, thereby relinquishing many of
the physiological guides normally used for accurate place-
ment of lesions. Narabayashi has reported improvement

in choreoathetosis in 78 percent of children with cerebral
palsy who underwent thalamotomy (41). He advocates
placing the lesion relatively anteriorly in the ventrolat-
eral thalamus (VOA-VOP) or just below the thalamus in
the field of Forel. In another series, Mundinger and asso-
ciates found that 18 percent of patients with choreoathe-
tosis were markedly improved and 32 percent moder-
ately improved at a late follow-up after thalamotomy
(36). All were at least moderately improved immediately
after surgery.

Broggi and coworkers have followed 33 patients with

cerebral palsy (27 congenital and 6 acquired) for one to
four years following thalamotomy (42). Lesions were
placed in the anterior portion of the ventrolateral thala-

310 / CHAPTER 14

mus (VOA, VOP, and zona incerta) in 32 of the patients

who had tremor and choreoathetosis. In 13 of these pa-
tients who had significant dystonia, a lesion was also

placed in the posterior thalamus (pulvinar medialis).

One patient received only the posterior lesion. All pa-
tients except one improved: 73 percent considerably and
24 percent moderately. Tremor and choreoathetosis

were most improved. Dystonia was less affected, and

spasticity was generally unaffected. Better results were
observed in younger patients. The results by Broggi and

associates are better than have been reported in most
other series. They attribute this largely to careful selec-
tion of surgical candidates.

Stereotactic destruction of a portion of the dentate nu-

cleus of the cerebellum is reported to improve
choreoathetotic movements in approximately one-half

of patients who undergo the procedure (43-45). Fairly

large and usually bilateral lesions are necessary. The ef-
fects of electrical stimulation are unclear; some studies
have shown improvement, but others have not (46,47).
This topic is discussed toward the end of this chapter, as

is sectioning of lumbar nerves for spasticity.

Stereotactic Technique and Considerations

To perform Stereotactic thalamotomy one needs (1) a

Stereotactic frame that provides rigid skull fixation and
accurate (± 1 mm) mechanical direction of an instru-
ment to the defined target, (2) a method by which radio-
logical data are processed to define the target point, (3)
an instrument with which to create the lesion, and (4) a
method to confirm the position of the Stereotactic instru-
ment intraoperatively.

Stereotactic Frames

The Leksell. the Todd-Wells, and the Brown-Roberts-

Wells systems are the Stereotactic instruments most
commonly used in the United States. Other systems in-
clude the Talairach system and the Reichert-Mundinger
system. A number of other devices, generally modifica-
tions of these five systems, are used by individual sur-
geons. Most of these systems have recently been modi-
fied for compatibility with computed tomography (CT),

magnetic resonance imaging (MRI). or both. The major

differences among instruments are in (1) the radio-
graphic methods used for locating the target point, (2)
the corrections for magnification and parallax on the in-
traoperative x-rays, and (3) the mechanism by which the
electrode carrier and the head are moved with respect to

each other.

In the Leksell system, after the target has been local-

ized on the x-ray films, the electrode carrier (which is on
a movable arc that can be thought of as lying on the
surface of a sphere) is moved with respect to the target so

that the center of the sphere lies at the target. In the

Todd-Wells system, the head is moved with respect to a

fixed arc to bring the target to the center of the sphere.
The center of the sphere lies at the crossing point of the
central ray of the anteroposterior and lateral x-ray

beams. With either of these systems, the superimposition

of the target on the center point of the electrode carrier
arc allows the target point to be approached through a
burr hole made at any point on the skull, as the center of
a sphere can be reached by any radius from the center to

its surface.

The following discussion of Stereotactic technique fo-

cuses on placement of a VIM lesion using intraoperative
plain radiographs and ventriculography to determine the
lesion coordinates and using physiological confirmation
of the final target site. However, many variations of tech-
nique exist. Some of these are noted in the following text,
but the full range of variations in technique is beyond the
scope of this chapter. A number of excellent surveys that

have recently been published provide a more extensive

description of Stereotactic methods (48-50).

Preoperative Preparation

The shape and size of the ventricular system, the de-

gree of cerebral atrophy, and the possible presence of
infarcts should be carefully evaluated with magnetic reso-
nance imaging in all patients. Any clotting or bleeding

disturbances must be detected and evaluated clinically

and with laboratory tests. Any medication that might
contribute to bleeding, such as aspirin, should be
stopped one to two weeks before surgery. The patient
should also be asked specifically about any history of
allergic reactions to iodinated contrast agents or to local
anesthetics. An antihistamine preparation, such as cime-
tidine (300 mg PO) and benadryl (25 mg PO), given at
bedtime and three hours prior to surgery will help pre-
vent allergic reactions. Dexamethasone (20 mg) given
intravenously approximately three hours before surgery
will further minimize allergic reactions. Phenytoin given
orally prior to surgery and intravenously on the morning
of surgery in a loading dose will reduce the possibility of
a seizure during ventriculography or postoperatively. An-
tibiotic prophylaxis, chosen with respect to the bacteria

prevalent in the hospital environment, should be given

one hour before surgery and for 24 hours following sur-

gery.

Thalamotomy is nearly always performed using local

anesthesia, as the patient's ability to note the effects of
electrical stimulation and to display his or her dyskinesia
is extremely helpful not only in confirming the proper

placement of the electrode but also in assessing the effect
of the lesion. Thus, the preoperative medications aim to

help the patient feel comfortable and alert. Hyperten-
sion, arrhythmias, nausea, and vomiting must be pre-

MOVEMENT DISORDERS / 311

vented: therefore, painful stimuli and reactions to ventri-
culography that could cause these disturbances must be
avoided. To achieve these goals, premedications, includ-
ing an atropine-like drug to prevent bradycardia, benzo-

diazapines and neuroleptic drugs to control anxiety, and
small doses of a narcotic to block painful stimuli, are
routinely given intravenously before the patient reaches
the operating room. The patient's alertness and vital
signs are monitored continuously.

It is also important that the patient can lie comfortably

for the duration of the operation. Most patients tolerate
reasonably well an operative time of approximately one
and one-half to two hours. For this reason, it is impor-
tant to perform the operation as efficiently as possible,
taking into account the need for careful radiological and
physiological determination of the target site. Therefore,
all of the necessary equipment should be assembled and
ready for use before the patient enters the operating
room. The alignment of the x-ray machine with the ster-

eotactic frame should be confirmed with test films. After
the patient enters the operating room, his or her hair is
shaved, and the scalp is prepared with an antiseptic. The
scalp is then infiltrated with a local anesthetic at the
points of attachment of the stereotactic frame, and the
frame is attached.

Determination of Lesion Coordinates

An intraoperative ventriculogram, CT scans, or a sagit-

tal MRI scan is used to determine the distance between
the anterior and posterior commissures (the AC-PC
line). Other points of reference (such as the foramen of
Monro or the pineal recess) may be used, but the AC-PC
plane seems to provide the most accurate coordinates for
placement of thalamic lesions. This measurement is
compared with a stereotactic atlas referenced to the AC-
PC plane, such as that prepared by Talairach and co-
workers (51) or by Schallenbrand and Bailey (52). The
coordinates for the target nuclear group or tract are cal-

culated. During surgery, electrophysiological testing is
used to check the lesion coordinates and to determine
the exact site for the lesion.

Some surgeons use CT scans to determine the lesion

coordinates and have reported good results (53,54).
Other centers calculate CT- or MRI-generated target
sites and then confirm and adjust these points with intra-
operative ventriculography (48). CT and MRI scans al-
low the internal capsule to be visualized, and this infor-
mation can assist in determining the target site. Because
scans are computer-generated images, they can be recon-
figured in various planes; also, overlays from a com-
puter-resident stereotactic atlas can be used to quickly
calculate the lesion coordinates (55).

CT-derived coordinates have been compared with

those calculated from ventriculography, with mixed re-

sults. Coordinates from slices 1.5 mm in thickness corre-
lated fairly well with the coordinates derived from ven-
triculography in one study, which found a correlation of
±0.67 mm for the x and y coordinates and ±2 mm for

the z coordinate (56). In another study. Birg and Mun-
dinger reported that CT-derived coordinates were accu-

rate to within ±0.6 mm of the final targets determined
physiologically (57). In yet another study, however, up to
8 mm discordance in the three-dimensional distance be-

tween the point determined by CT and the point deter-
mined by ventriculography was observed (58). Although
the coordinates derived from ventriculography are gener-
ally thought to be more accurate, this is not altogether
clear. Some investigators believe that introduction of the
contrast material may distort the ventricle and nearby
anatomical relationships, and they note that ventricu-
lography itself adds to the risk and overall morbidity of
the procedure (54,55).

Magnetic resonance imaging may be the imaging

method by which functional neurosurgery is performed
in the future. MRI defines the commissures, thalamic
nuclei, and individual variations in anatomy better than
any method in current use. However, at present there is
some concern that anatomical structures may not be dis-

played in a linear manner but may be subject to some
distortion by the magnetic field (54). The ideal imaging
and surgical method would allow intraoperative visual-
ization of the thalamic nuclei. A computer would then
create three-dimensional reconstructions based on these
radiological data and would plot the ideal lesion point
and the best trajectory to approach that point based on
the individual patient's data, brain-atlas data, and a da-
tabase from other thalamotomy patients.

At the present time, most centers determine the lesion

coordinates using ventriculography and physiological
confirmation (27,59). For determination of the initial
coordinates, intraoperative anteroposterior and lateral
x-rays are taken with the patient's head in the frame. The
approximate position of the target point in the individ-
ual patient's brain can at this point be determined to
within 5 to 10 mm by using an overlay of the outline of

the floor of the anterior fossa, sella, and middle fossa; the
outline also shows the position of the third and lateral
ventricles and the anterior and posterior commissures.
This overlay mask is constructed from ventriculograms
made at the same tube-patient-film distances. With a
Todd-Wells type of frame, an initial adjustment of the
position of the head within the device is made to bring
the target point, determined from the outline mask, to
the center of the electrode arc.

There are two stereotactic approaches to the ventro-

lateral thalamus: one frontal, through a burr hole placed

1 to 2 cm in front of the coronal suture, and the other

parietooccipital, through a posteriorly placed burr hole.
Laitinen surveyed 16 neurosurgeons about their pre-
ferred approach for the placement of a thalamic lesion.

312

CHAPTER 14

He found that 15 preferred the frontal approach,

whereas only one preferred the parietooccipital ap-

proach (60). The frontal approach places the lesion-mak-

ing electrode along the long axis of VIM and creates a

favorable situation for making a radiofrequency lesion

conforming to the shape of the nucleus. In addition, the
lateral ventricle is most easily and safely cannulated for
ventriculography through the frontal approach. Patients
are most safely operated on in the supine position in
terms of cardiovascular stability and comfort, and the

design of the commonly used stereotactic instruments
makes a frontal trajectory easier to achieve than a poste-
rior trajectory when the patient is supine. The posterior
trajectory has the advantage of allowing an exploring mi-
croelectrode to pass through the ventralis caudalis (Vci-
Vce) sensory relay nucleus prior to entering VIM and to

record thumb, index finger, and buccal commissure neu-

rons, which will confirm the correctness of the lateral
coordinate for the target neurons in VIM (61).

For a frontal approach, the site for the burr hole is

approximately 12 cm posterior to the nasion and 2.5 cm
from the midline. The scalp over the burr hole site is

anesthetized, the burr hole made, the dura opened, the
crown of a gyrus cauterized lightly, and the ventricle can-
nulated. Ventriculography is then performed in order to
visualize the outline of the third ventricle and the ante-
rior and posterior commissures. At present, nonionic io-
dinated water-soluble contrast medium, nearly isotonic

with cerebrospinal fluid, is generally used, in a volume of

5 cc injected at a rate of 1 cc per second. Anteroposterior

and lateral radiographs are taken immediately after the
injection is completed.

A line is drawn on the ventriculogram between the

anterior and posterior commissures (Fig. 3), and the tar-

get point is calculated with reference to the coordinates
given in a stereotactic atlas. A system of Cartesian coordi-
nates is used in the atlas to designate the position of
various structures in relation to the AC-PC line. In rela-
tion to a 25 to 26 mm AC-PC line, a lesion in the VIM
nucleus lies approximately 5 mm posterior to the mid-
point of the AC-PC line, 1 mm above the line, and 13
mm lateral to the middle of the third ventricle (27t62).
These coordinates assume that a frontal trajectory'will

be used for the electrode insertion. The zero point for the
coordinate system is the midpoint of the AC-PC line.

Various methods have been devised for correcting co-

ordinates to compensate for variations in the size of the

thalamus. The simplest technique is to apply a propor-
tionate correction based upon the ratio of the AC-PC
line in the patient to the coordinates given in an atlas.
The actual site of the lesion may vary several millimeters
from the initial target coordinates, depending upon the
physiological testing.

Once the coordinates are calculated, the head is

moved again to bring the target point to the center of the
electrode arc. A third set of anteroposterior and lateral

x-rays is taken to confirm the position.

Physiological Confirmation

Because of the variation in the size and shape of the

thalamus in individual patients, the coordinates derived

from the atlas, even though they have been adjusted with

B

FIG. 3. (A) Lateral positive contrast ventriculogram showing the outline of the third ventricle and the
anterior and posterior commissures. (B) Tracing of A on which the intercommissural (AC-PC) line and the
target point in VIM are constructed. The long axis of VIM and the electrode trajectory are also indicated.
(From reference 27, with permission.)

MOVEMENT DISORDERS / 313

reference to the length of the patient's AC-PC line, must
be further refined by physiological testing. Kelly, Der-

ome, and Guiot determined that the anterior border of

the Vci and Vce, as determined from stereotactic atlases,
varied as much as 6 mm in location when determined
electrophysiologically (63). They found that it was neces-
sary to move the lesioning probe from the radiographi-
cally determined point to a new site in 71 of 100 patients.
Most of these adjustments required moving the electrode
more laterally.

The electrode is slowly advanced to the preselected

coordinates, and the placement of the electrode is evalu-
ated by noting the patient's response to stimuli adminis-
tered through the electrode or by stimulating the patient

tactilely and recording the response, or by using both

methods. The most valuable confirmatory sign is the pa-

tient's reporting, in response to stimulus, of paresthesias
in the contralateral thumb and index finger and in the
corner of the mouth. The target lies immediately ante-

rior to the sensory relay cells in the ventralis caudalis

(Vce-Vci) for these areas of the hand and face (8). Kelly
and associates found, in a study of 97 patients, that when
the lesion was made in the area of the somatic representa-

tion of the thumb, buccal commissure, and index finger

in the thalamus, complete abolition of the tremor oc-
curred (63).

The stimulus intensity required to evoke the sensory

responses can also indicate whether the electrode is at the
target site or is posterior or anterior to it. With a 2.1-mm
diameter electrode as the negative pole and the positive

electrode on the cut edge of the scalp, and with square-

wave pulses of 1-ms duration at 20 Hz, paresthesis
should be produced at an intensity of 3 to 4 volts if the
electrode is in the VIM nucleus. A lower threshold sug-
gests that the electrode is in the VC; a higher threshold
suggests that the electrode is anterior, lateral, or medial
to the VIM. Low-frequency electrical stimulation at the
target site in the VIM frequently accentuates the pa-

tient's tremor, whereas high-frequency stimulation will
abolish it (48).

VIM localization can also be confirmed by using a

microelectrode to record the tremor-linked discharges of
VIM neurons and discharges in response to such kines-
thetic stimuli as muscle stretch. VC neurons can be iden-

tified by recording their discharge to stimulation of the

hand and face (2,62). After the recording, the microelec-

trode is replaced with the lesion-making electrode. This
technique can make localization more precise, but it also
prolongs the procedure, which in some cases is not desir-
able.

Generation of Lesion

A radiofrequency lesion whose size and rate of devel-

opment are controlled by monitoring the temperature at

the electrode tip is the method most commonly used by

stereotactic surgeons in the United States. The size of the
lesion depends on the diameter of the electrode, the

length of the uninsulated tip, and the temperature mea-

sured at the electrode tip (64,65). The lesion has the
shape of an oblate spheroid, or a football. A radiofre-
quency lesion made in VIM with a 2.1-mm diameter
electrode having a 5-mm bare tip and a tip temperature-^
maintained at 75° for 60 seconds will generally produce

excellent control of tremor with minimal side effects.

Lesions can also be made between smaller diameter elec-

trodes positioned several millimeters apart. The leuko-
tome can make a unidirectional lesion, but it is more
likely to cause a hemorrhage. Cryogenic lesions require
insertion of a probe several times larger in diameter than
an electrode, and, unless the probe size can be reduced,
widespread readoption of this technique is unlikely.

When the lesion is made, the patient is asked to hold

up the contralateral arm and to open and close the hand
in a fist to test for the possible occurrence of weakness.

The patient is then tested for the presence of residual
tremor; if necessary, the lesion is then enlarged or addi-
tional lesions are made in adjacent tissue.

Complications and Postoperative Care

Surgical complications occur in 1 to 2 percent of pa-

tients. The most common complications are infection,

seizures, or subdural hematoma. Subdural hematoma is
more likely in patients with considerable cortical atro-
phy. Cortical atrophy can also lead to an accumulation

of air in the subdural space over the cerebral hemi-

spheres during surgery, which can produce postoperative
fever and confusion. Specific neurological deficits re-
lated to the size of the lesion and the accuracy of its
placement can also occur. Paresthesias and diminished
sensation in the contralateral face, index finger, and

thumb can occur after a VIM lesion because of extension
of the lesion into the Vci-Vce complex, which lies imme-
diately posterior to the VIM. Contralateral hemiparesis
can occur owing to extension of the lesion laterally into
the corticospinal tract of the posterior limb of the inter-
nal capsule, which lies immediately lateral to the VIM
nucleus. Lesions located too inferiorly risk causing he-

miballismus, and lesions too medial, especially if they

are bilateral, can be associated with memory distur-
bances. Language deficit can occur with a unilateral left
thalamic lesion. The mechanism of this deficit is not
completely understood.

Bilateral lesions are more likely to produce a speech

deficit than are unilateral lesions. Bilateral thalamic le-
sions can markedly decrease voice volume, a pseudobul-
bar type of deficit that has been reported to occur in
approximately 30 percent of patients with bilateral le-

sions. For this reason, the benefits and risks of bilateral

314 CHAPTER 14

lesions must be weighed especially carefully in the indi-
vidual patient. If a lesion contralateral to the first lesion
is desired, a period of approximately six months should

generally be allowed to pass before the second operation

in order to assess the effects of the initial lesion properly.

Following surgery, the patient should be observed in

an intensive care unit for one to two days. Most patients
exhibit a mild contralateral central facial weakness for
several days and may exhibit mild hypotonia and ataxia
in the contralateral limbs. There may be a tendency to
neglect the left hand for several weeks after right thala-
mic lesioning. It is important to explain to the patient
that, although the tremor may be completely abolished,
it will usually take a few weeks to several months to re-
learn to use the hand for skilled activities if the tremor
had previously rendered the hand nonfunctional.

Anticonvulsant medication is discontinued at the

time of discharge from the hospital. The requirement for

antiparkinsonism medication for rigidity or bradykine-

sia may be reduced in some patients.

OTHER PROCEDURES

Nerve-Sectioning Procedures

Sectioning of nerves is done primarily for torticollis,

although it has also been used to relieve spasticity in
persons with cerebral palsy. Because torticollis, unlike

the other movement disorders, is confined specifically to

one area of the body (the head and neck), selective sec-
tioning of the anterior roots of the cervical nerves (and
sometimes the spinal accessory) can be used to denervate
partially the head and neck muscles involved, thereby
relaxing the muscles and allowing the head to return to a
more normal position. The primary side effect is a loss of
tone in the neck musculature if extensive bilateral dener-
vation is performed. This can result in some difficulty in

holding the head erect and in swallowing.

The initial step in treating torticollis by denervation of

the involved muscles is to characterize the patient's head
and neck movement and the muscular contractions
causing it. It is important to identify the direction and
degree of rotation, flexion, extension, and lateral tilt of
the head. Electromyography is used to analyze the pat-
tern of muscle contraction. A common pattern in torti-
collis is to find contraction of one sternocleidomastoid
muscle and the contralateral splenius capitus, with vary-
ing degrees of contraction of the contralateral levator
scapulae and trapezius muscles (66). Nerve blocks with
local anesthetic can be used to identify nerve roots of the
muscles involved in producing the torticollis. The spinal
accessory nerve can be blocked either before it enters the
sternocleidomastoid or lower in the neck, before it enters

the trapezius. The posterior rami of the anterior roots of

the upper cervical nerves can be blocked as they exit the
intervertebral foramina.

The primary surgical decision to be made is whether to

offer the patient a myotomy or a nerve-sectioning proce-
dure as the initial operation. Myotomy has the benefits
of low risk and a short hospital stay, but it is most success-
ful in comparatively mild cases. The primary muscles
sectioned are the sternocleidomastoid and the contralat-
eral splenius capitus (66). The levator scapulae and the
upper portion of the trapezius can also be sectioned. My-
otomy usually produces about a 50 percent improve-
ment in the resting position of the head. Improvement

generally occurs over one to three months. Physical ther-

apy is important to achieve and to maintain the improve-\
ment.

Sectioning the nerves to the involved muscles is gener-

ally more effective than myotomy. Nerve-sectioning

procedures for torticollis can be divided into two types:
(1) extraspinal section of the dorsal rami of the anterior

roots (39), and (2) intraspinal sectioning of the anterior

roots, which can be combined with intraspinal section-
ing of the spinal accessory nerve (40). Extraspinal sec-
tioning of the dorsal rami is advantageous because it

does not require that the dura be opened and does not

denervate the anterior neck musculature. The dorsal
rami can be sectioned down to the C7 vertebra. Bertrand
and associates reported on 35 patients with spasmotic
torticollis who were treated with selective denervation.
In most instances, the posterior rami of C1 to C4 were
done on one side and the posterior rami of C1 to C5 on
the other. Depending on the individual patient, the spi-

nal accessory was sometimes also sectioned. In this se-

ries, 34 percent had an excellent result, 54 percent a good
result, 9 percent a fair result, and 1 percent a poor re-
sult (39).

Intraspinal sectioning of the anterior roots and the spi-

nal accessory nerve can be performed with low morbid-
ity using microsurgical techniques. The three major de-
cisions to be made are (1) how many anterior roots to
section, (2) whether to do a unilateral or bilateral proce-
dure, and (3) whether to section the spinal accessory
nerve. Sectioning of C4 will result in paralysis of the ipsi-

lateral diaphragm, and bilateral sectioning of the C1 to
C3 roots can result in difficulty holding the head erect
and swallowing (40). For these reasons, a combined tech-

nique consisting of rhizotomies of the anterior roots of

C1 to C3 (to denervate the posterior cervical muscles on
the side of the head rotation and tilt) and of contralateral
sternocleidomastoid muscle myotomy has been advo-
cated (27). Fixed head and neck deviation can be re-

turned to about 10° to 20° from the midline in many

patients by this approach.

A lumbar nerve-sectioning procedure may be useful in

reducing spasticity in patients with cerebral palsy. Pea-
cock, Arens, and Berman performed selective posterior
rhizotomies on 60 children with cerebral palsy (67). The

MOVEMENT DISORDERS / 315

procedure involves stimulating the rootlets of the poste-
rior L2 to S1 nerves and sectioning the rootlets that dem-
onstrate an abnormal motor response. Patients were fol-
lowed one to five years postoperatively. All patients
except one improved following surgery. Children whose
primary disability was spasticity of the lower extremity
improved most, and those with severe athetosis or
marked contractures improved least. When surgery was

combined with intensive physical and occupational post-
operative therapy, functional improvement was dra-
matic in some instances.

those reported earlier by Waltz and Pani when a conven-

tional two-electrode system was used (73). Groen and
associates have reported an excellent result in the treat-
ment of a patient with severe myoclonic dystonia by us-
ing chronic stimulation with epidural electrodes im-
planted over the upper cervical cord (74).

The mechanism for the improvement noted in some

studies following chronic electrical stimulation is not un-
derstood, and the results have been inconsistent. For

these reasons, few patients are now being treated with
these methods.

Chronic Electrical Stimulation

The results of chronic electrical stimulation for the

treatment of movement disorders have been inconsis-
tent. In one study, chronic electrical stimulation of the
upper cervical spinal cord produced relief from torticol-
lis in approximately one-third of patients (68); but in
another study it was ineffective for the torticollis, al-
though it ameliorated the cervical pain that may accom-
pany torticollis (5). Cooper and associates reported on 50

patients in whom they inserted electrodes bilaterally

over the anterior cerebellar lobes for the treatment of
cerebral palsy (69). Patients were rated in eight func-
tional areas preoperatively and one and five months
after the initiation of cerebellar stimulation. Stimulation
parameters varied somewhat among patients, but a com-
monly used protocol was stimulation at 10 volts, 200
cycles per minute, administered at eight-minute inter-
vals. These researchers noted improvement in spasticity

as well as in athetosis, speech, and functional status. In a
slightly later report, however, Cooper's group noted that
the overall improvement appeared to be less than ini-

tially observed (70). Penn and associates performed two

studies—a prospective study and a double-blind study—
of cerebellar stimulation for cerebral palsy (71). In the
prospective study, they found that 11 of the 14 patients
(79 percent) were better functionally after 1 to 44
months of stimulation. However, in the double-blind
study, no significant changes in motor function could be

observed.

Waltz, Reynolds, and Riklan designed a four-elec-

trode system that they implanted epidurally over the up-

per cervical spinal cord (C2 to C4) in patients with
various movement disorders. In a series of 160 patients,
improvement of some degree was noted in 84 percent of
the 75 patients with cerebral palsy, 67 percent of the 42
patients with dystonia, 62 percent of the 21 patients with

torticollis, and 73 percent of the 22 patients with post-
traumatic neurologic loss (72). In patients with cerebral

palsy, spasticity was decreased both objectively and sub-

jectively in 94 percent; athetosis was improved in 67 per-

cent of the patients; and posture and balance improved
in 83 percent. These results are considerably better than

Neurotransmitter Augmentation

Pumps

Relatively little research has been done into the possi-

bility of delivering neurotransmitters or medications for
the treatment of movement disorders directly into the
cerebrospinal fluid or even into the parenchyma of the
brain. Such systems have been used, however, with
mixed results, to deliver analgesics for chronic pain, med-
ications such as methotrexate for central nervous system

neoplasms, and cholinergic drugs for Alzheimer's dis-

ease. The potential advantages of intrathecal drug admin-
istration include bypassing the problems of systemic side
effects, peripheral drug inactivation, poor drug absorp-
tion, serum protein binding, inadequate blood-brain
penetration, and poor patient compliance. In parkinson-
ian patients, inconsistent delivery of levodopa or other
dopaminergic drugs to the striatum may account for the
"on-off" response; a continuous pump infusion might

alleviate this problem. At the present time, however, in-

trathecal drug administration has largely been ineffec-
tive in the treatment of Parkinson's disease, primarily
because of the development of toxicity or severe psychiat-
ric disturbances (75). However, success with intrathecal
administration of baclofen for the treatment of spasticity

has been reported. Ochs and associates followed the

progress of 28 patients with severe, intractable spasticity

for up to two years who were treated by chronic intrathe-

cal administration of baclofen using a programmable
drug-administration device. Infusion of 50 to 800 mg per
day of baclofen abolished spasticity with few complica-
tions or side effects (76).

Transplantation

The transplantation of dopaminergic adrenal medul-

lary tissue into the basal ganglia of patients with intract-

able Parkinson's disease has produced inconclusive re-
sults (77-79). A recent survey of worldwide experience
in 138 patients indicated that 52 percent were at least
moderately better more than three months after the adre-
nal implant (80). At the present time it appears that

316 / CHAPTER 14

younger patients respond better and with fewer compli-
cations than older patients. The most consistent im-
provement in motor performance has been an ameliora-

tion of gait difficulty and postural instability;

antiparkinsonian medications can be reduced in some
patients. The mechanism for this modest improvement
is not known. Although it was first proposed that the
grafts would provide a source of dopamine, there is
currently little evidence to suggest that the implants re-
lease dopamine. Pathological study of three brains of pa-

tients who died from unrelated causes following trans-
plantation showed no evidence of surviving adrenal
medullary tissue (79,81,82).

REFERENCES

1. Kelly PJ. Gillingham FJ. The long term results of stereotactic sur-

gery and L-dopa in patients with Parkinson's disease. A 10 year
follow-up siudy.J Neurosurg 1980;53:332-337.

2. Nagaseki Y, Shibazaki T, Hirai T, et al. Long-term follow-up re-

sults of selective VIM-thalamotomy. •] Neurosurg 1986;65:296-
302.

_3. Rajput AH, Offord KP, Beard M, Kurland LT. Epidemiology of

parkinsonism: incidence, classification and mortality. Ann Neural

16:1984;278-282.

4. Schoenberg BS. Epidemiology of movement disorders. In: Mars-

den CD, Fahn S, eds. Movements disorders 2. London: Butter-
worths, 1987; 17-32.

5. Ojemann GA, Ward AA Jr. Abnormal movement disorders. In:

Youmans JR, ed. Neurological surgery: a comprehensive reference
guide to the diagnosis and management of neurosurgical problems,
3rd ed. Philadelphia: WB Saunders, 1990:4227-4262.

6. Scherman D, Desnos C, Darchen F, Pollak P. Javoy-Agid F, And

Y. Striatal dopamine deficiency in Parkinson's disease: role of ag-
ing. Ann Neural 1989;26:551-557.

7. German DC, Manaye K, Smith WK, Woodward DJ, Saper CB.

Midbrain dopaminergic cell loss in Parkinson's disease: computer
visualization. Ann Neural 1989;26:507-514.

8. Kelly PJ. Applications and methodology for contemporary stereo-

tactic surgery. Neural Res 1986;8:2-12.

9. Hoehn MH, Crowley TJ, Rutledge CO. Dopamine correlates of

neurological and psychological status in untreated parkinsonism. J
Neural Neurosurg Psychiatry 39:1976;941 -951.

10. Bucy P. The surgical treatment of extrapyramidal disease. J Neural

Neurosurg Psychiatry 1951;14:108-117.

11. Meyers R. The modification of alternating tremor, rigidity and

festination by surgery of the basal ganglia. Res PublAssoc Res Nerv
MentDis 1942;21:602-665.

12. Cooper IS. Involuntary movement disorders. New York: Harper

and Row, 1969.

13. Bertrand C. Localization of lesions. J Neurosurg \ 966;24:446-448.
14. Narabayashi H, Yokochi F, Nakajima Y. Levodopa-induced dys-

kinesias and thalamotomy. J Neural Neurosurg Psychiatry

1984;471:831-839.

15. Hardy TL, Bertrand G, Thompson CJ. Thalamic recordings dur-

ing stereotactic surgery. I. Surgery topography of evoked and none-

voked rhythmic cellular activity. Appl Neurophysiol 1979;42:185-

197.

16. Kaplan RJ. Movement disorders. In: Rosenberg RN, Grossman

RG, eds. The clinical neurosciences, vol 2. New York: Churchill
Livingstone, 1983:1325-1348.

17. Hassler R, Mundinger F, Riechert T. Stereotaxis inparkinson syn-

drome: clinical-anatomical contributions to its pathophysiology:

with an atlas of the basal ganglia in parkinsonism by R. Hassler.

New York: Springer-Verlag, 1979.

18. Stern J. Ward A Jr. Inhibition of the muscle spindle discharge by

ventrolateral thalamic stimulation: its relation to parkinsonism.
Arch Neural 1960;3:193-204.

19. Stern J, Ward AA Jr. Supraspinal and drug modulation of the

alpha motor system. Arch Neural 1962;6:404-413.

20. Stern J, Ward AA Jr. The relationship of the alpha and gamma

motor systems to the efficacy of the surgical therapy of parkinson-

ism. J Neurosurg 1963:20:185-187.

21. Bader E. Extrapyramidal symptoms following thorazine medica-

tion. IllMedJ 1956;109:28-29.

22. Struppler A, Burg D, Lucking CH, Velho F. The mode of innerva-

tion following thalamotomy and subthalamotomy. Confin Neurol

1974;36:347-354.

23. Yanagisawa N, Narabayashi H, Shimazu H. Thalamic influences

on the gamma motor system. Arch Neurol 1963;9:348-357.

24. Lieberman JS, Copack PB, Oilman S. Cryogenic lesions in the

ventrolateral nucleus and pulvinar. Arch Neurol 1974;30:375-382.

25. Barbeau A, McDowell F. L-dopa and parkinsonism. Philadelphia:

FADavis, 1970.

26. Barbeau A. The clinical physiology of side effects in long term

L-dopa therapy. Adv Neurol 1974;5:347-365.

27. Grossman RG. Surgery for movement disorders. In: Jankovic J,

Tolosa E, eds. Parkinson's disease and movement disorders. Balti-

more-Munich: Urban & Schwarzenberg 1988:461-470.

28. Gillingham PL. Bilateral stereotactic lesions in the management of

parkinsonism. J Neurosurg 1966;24:449-453.

29. Cooper IS, Riklan M, Stellar S, et al. A multidisciplinary investiga-

tion of neurosurgical rehabilitation in bilateral parkinsonism. J
Am GeriatrSac 1968;16:1177-1306.

30. Benadid AL, Pollak P, Louveau A, Henry S, de Rougemont J.

Combined (thalamotomy and stimulation) stereotactic surgery of
the VIM thalamic nucleus for bilateral Parkinson's disease. Appl
Neurophysiol 1987;50:344-346.

31. Riechert T. Stereotactic brain operations: methods, clinical

aspects, indications. Bern: Hans Huber, 1980.

32. Andrew J, Fowler CJ, Harrison MJG. Tremor after head injury

and its treatment by stereotactic surgery. J Neurol Neurosurg Psy-
chiatry 1982;45:815-819.

33. Bullard DE, Nashold BS Jr. Stereotactic thalamotomy for treat-

ment of posttraumatic movement disorders. J Neurosurg

1984;61:316-321.

34. Krayenbuhl H, Yasargil M. Relief of intention tremor due to multi-

ple sclerosis by stereotactic thalamotomy. Confin Neurol

1962;22:368-374.

35. McDowell FH, Cedarbaum JM. The extrapyramidal system and

disorders of movement. In: Baker AB, Baker LH, eds. Clinical
Neurology, vol 3. Philadelphia: JB Lippincott. 1988;l-97.

36. Mundinger F, Reichert T, Disselhoff J. Long term results of stereo-

tactic operations on extrapyramidal hyperkinesias (excluding par-
kinsonism). Confin Neurol 1970;32:71-78.

37. Andrew J, Fowler CJ, Harrison MJG. Stereotaxic thalamotomy in

55 cases of dystonia. Brain 1983;106:981-1000.

38. Stell R, Thompson PD, Marsden CD. Botulinum toxin in spasmo-

tic torticollis. J Neurol Neurosurg Psychiatry 1988;51:920-923.

39. Bertrand C, Molina-Negro P, Martinez S. Technical aspects of se-

lective denervation for spasmodic torticollis. Appl Neurophysiol

1982:45:326-330.

40. Hamby WB, Schiffer S. Spasmodic torticollis: results after cervical

rhizotomy in 50 cases. J Neurosurg 1969;31:323-326.

41. Narabayashi H. Stereotaxic surgery for athetosis of the spastic state

of cerebral palsy. Confin Neurol 1962;22:364-367.

42. Broggi G, Angelini L, Bono R, Giorgi C, Nardocci N, Franzini A.

Long term results of stereotactic thalamotomy for cerebral palsy.

Neurosurgery 1983;12:195-202.

43. Nashold B, Slaughter DG. Effects of stimulating or destroying the

deep cerebellar regions in man. J Neurosurg 1969;31:172-186.

44. Siegfried J, Esslen E, Gretner U, Ketz E, Perret E. Functional anat-

omy of the dentate nucleus in the light of stereotactic operations.
Confin Neurol 1970;32:l-10.

45. Gornall P, Hitchcock E, Kirkland IS. Stereotaxic neurosurgery in

the management of cerebral palsy. Dev Med Child Neurol

1975;17:279-286.

46. Cooper I, Riklan M, Amin I, Waltz J, Cullinan T. Chronic cerebel-

lar stimulation in cerebral palsy. Neurology 1976;26:744-753.

47. Penn R, Myklebust B, Gottlieb G, Agarwak G, Etzel M. Chronic

cerebellar stimulation for cerebral palsy: prospective and double-
blind studies. J Neurosurg 1980;53:160-165.